Time-Optimal Control by Fuzzy Controller using Evolution Strategy

Abstract

Time-optimal control by fuzzy controller is considered in this paper. Usually, very complicated numerical computation is necessary to obtain the solutions of time-optimal control problems. In this study, it is shown that such problems can be solved easily by fuzzy controllers obtained by using the Evolution Strategy(ES)as the learning algorithm of the parameters of Gaussian membership functions. Especially, this proposed global optimization approach is shown to be very useful and promising for the highly nonlinear problems which are very difficult to be solved by the conventional gradient based approaches such as a steepest descent, conjugate gradient, quasi-Newton methods.